Initial model upload

README.md

@@ -27,7 +27,7 @@ We validate the execution of DeepSeek-R1 model with a Triton-based operator libr
 
 We use a variety of Triton-implemented operation kernels  to run the DeepSeek-R1 model. These kernels come from two main sources:
 
-- Most Triton kernels are provided by FlagGems (https://github.com/FlagOpen/FlagGems). You can enable FlagGems kernels by setting the environment variable USE_FLAGGEMS. 
+- Most Triton kernels are provided by FlagGems (https://github.com/FlagOpen/FlagGems). You can enable FlagGems kernels by setting the environment variable USE_FLAGGEMS.
 
 - Also included are Triton kernels from vLLM, such as fused MoE.
 
@@ -38,10 +38,10 @@ We use a variety of Triton-implemented operation kernels  to run the DeepSeek-R1
 | Basic Image | basic software environment that supports FlagOS model running | <IMAGE>             |
 # Evaluation Results
 
-## Benchmark Result 
+## Benchmark Result
 
-| Metrics            | DeepSeek-R1-H100-CUDA | DeepSeek-R1-FlagOS-metax |
-|:-------------------|--------------------------|-----------------------------|
+| Metrics           | DeepSeek-R1-H100-CUDA | DeepSeek-R1-FlagOS-metax |
+|-------------------|--------------------------|-----------------------------|
 | cmmmu | 49.11 | 42.89 |
 | mmmu | 57.44 | 47.56 |
 | mmmu_pro_standard | 38.4 | 30.21 |